Update to support OpenShift GitOps 1.11

[sabre1041]

Update to support OpenShift GitOps 1.11

[gnunn1]

LGTM however I'll leave to the others to comment on changing patch-channel to patch-subscription since everywhere else AFAIK we are using patch-channel. Or maybe an oppty to start moving this into inline patches in kustomization.yaml?

[sabre1041]

LGTM however I'll leave to the others to comment on changing patch-channel to patch-subscription since everywhere else AFAIK we are using patch-channel. Or maybe an oppty to start moving this into inline patches in kustomization.yaml?

We need to patch more than the channel starting in 1.11 since it deploys to a unique namespace

[pittar]

@gnunn1 , I'd be happy to start moving these patches inline. Maybe I can find some time over the holidays :)

[pittar]

@sabre1041 , if we moved the patch inline, would that solve the naming issue?

[sabre1041]

@sabre1041 , if we moved the patch inline, would that solve the naming issue?

I do not understand the the issue with changing the name of the patch. It is performing more than just patching the channel

[gnunn1]

I'm ok with the name change given @sabre1041 comment.

With the change in namespace do we need a readme comment about updating from other overlay versions as the namespace change will break things if you don't have pruning enabled in Argo CD?

[sabre1041]

I'm ok with the name change given @sabre1041 comment.

With the change in namespace do we need a readme comment about updating from other overlay versions as the namespace change will break things if you don't have pruning enabled in Argo CD?

So doing more research. The use of a separate namespace began with 1.10. Maybe we move forward with the following:

1. Apply the new component to the 1.10 overlay
   1. This is challenging if this overlay is in use
2. Add notes in overlay README

Open for discussing options

[adetalhouet]

There is a typo in the readme; beside that, look good to me.
Moving the operator to its own namespace is a good idea (I did this a year ago and it removed a lot of OLM related issues).

[adetalhouet]

I'll let @gnunn1 or @pittar merge this given the outstanding discussion. From my standpoint, it is good to go.

[pittar]

I'd never argue against @sabre1041 ;)